Last week, YouTube CEO Susan Wojcicki wrote a blog post in which she talked about the platform’s commitment to leaving up controversial videos even when they are offensive. This week, the company posted a new message about the videos they have decided to take down — and, YouTube says, it’s taking down many more videos than it ever has before.
Julia Alexander sums it up at The Verge:
YouTube’s teams have removed more than 100,000 videos and 17,000 channels since the company implemented changes to its hateful content policies in June.
Those numbers are approximately five times as many than the company’s last quarter, according to a new blog post from YouTube about the company’s attempts to tackle a growing number of hateful and dangerous videos on the platform. This also includes doubling the removal of comments (more than 500 million) that were found to be hateful. Some of these channels, videos, and comments are old and were terminated following the policy change, according to the blog post. This could account for the spike in removal numbers.
As usual when tech companies announce platform-scale statistics, the numbers involved can boggle the mind.
When you hear that YouTube removed 100,000 videos, does that sound like a lot, or a little?
Do the 17,000 removed channels represent the core of YouTube’s problems with extremism and hate speech, or are they the tip of the iceberg?
Does YouTube removing 500 million comments indicate a spectacular advance in moderation technology, or had it previously been ignoring a lot of low-hanging fruit?
These questions are all but impossible to answer, because as outsiders we have very little knowledge of what’s on the platform. We find out about enforcement actions only after the fact. In the meantime, we search, we see what’s trending, and we speculate — some of us in good faith, and others not.
But after a summer of cascading PR crises, YouTube is keen to convey the sense of a steady hand at the wheel. Today’s blog post is the first of a planned four-part series on “responsibility,” which the company has divided into four (other) R’s. (Coming up after “remove” are “raise,” “reward,” and “reduce.”)
I can’t help but feel like there’s a fifth R missing from this list: “review,” maybe, or “recourse.” A common theme in YouTube’s bad summer has been that human beings caught up in the platform’s machinery often have no good mechanism for appeal.
What if you’re a creator who has become the target of a bully who inspires his followers to harass you?
What if YouTube is algorithmically promoting videos of your children to pedophiles?
What if your videos on LGBT themes are age-gated and demonetized?
What if your videos about racism get caught up in an anti-racism sweep?
If you’re a creator, you can file an appeal — which, for all but the largest accounts, means filling out an online form and praying. (Some LGBT creators are suing YouTube over demonetization issues, but the case seems likely to advance very far.)
At this stage in its evolution, YouTube in some ways resembles a nation-state. But it lacks one of a state’s most essential features: a legitimate justice system. There is almost no way in which, on a decision-by-decision basis, the hard-working folks at YouTube are truly accountable. And it is for that reason that, no matter how many times YouTube tells us that it has removed a large number of videos from the site, its enforcement decisions will struggle to convey a sense of their legitimacy.
Facebook’s answer to this issue is its still-forthcoming independent oversight board, which will attempt to devolve some of its power to a group of outsiders. In recent discussions with Googlers, I’ve learned that the company is watching its rival’s effort closely — but has stopped well short of spinning up a similar program.
There are lots of reasons for Google to let Facebook take the lead here, of course. But it strikes me that before you decide to give away some of your power to outsiders, you first have to come to terms with how powerful you are. And I wonder, at this late date in 2019, whether YouTube truly has.
Today I’m excited to introduce The Ratio, an Interface feature that attempts to keep track of how the day’s news could affect public perception of our biggest tech companies. Some items will here effectively serve as previews of what’s to come later in the newsletter; other times, what you see here will be all I have to say on the subject. My hope is that this section complements the day’s lead column by giving you an at-a-glance look at the day’s other big news.
↗️ Trending up: Facebook will no longer scan users’ faces by default, a win for privacy advocates. It’s also preventing the Department of Homeland Security from making fake accounts as part of its investigations.
↘️ Trending down: Twitter CEO Jack Dorsey’s account got hacked on Friday, and the hackers tweeted racist garbage for an uncomfortably long time. Bad for public confidence in Twitter — and scary when you imagine something similar happening to the president’s Twitter account.
Trending sideways: Google is on the verge of settling an FTC investigation for cheap — but is about to get hit with an antitrust investigation by US attorneys general.
Read on for details.
⭐ More than half of state attorneys general are cooperating on an investigation into Google over antitrust issues. They are expected to announce the investigation at a news conference on Monday in Washington. (Tony Romm / Washington Post)
It is unclear whether some or all of the attorneys general also plan to open or announce additional probes into other tech giants, including Amazon and Facebook, which have faced similar U.S. antitrust scrutiny.
Over the past year, regulators around the country have grown increasingly wary regarding the power wielded by Silicon Valley, questioning whether the industry’s access to vast amounts of proprietary data — and deep pockets — allow companies to gobble up rivals and maintain their dominance to the detriment of consumers. Two federal antitrust agencies have opened probes targeting the industry broadly, while lawmakers in Congress have grilled executives from Amazon, Apple, Facebook and Google about the business practices.
But Google is getting off easy with the FTC. The company will pay between $150 and $200 million to end an investigation into YouTube over alleged violations of a children’s privacy law. I’m hearing the settlement will be announced Wednesday. (Margaret Harding McGill / Politico)
Google is indirectly providing services to US Citizenship and Immigration Services (USCIS), a branch of the Department of Homeland Security. The finding is likely to rankle employees. (Colin Lecher / The Verge)
⭐ Facebook’s internal review of its efforts to fight interference in the Brazilian election found serious flaws. The review stands in contradiction to some of the company’s public statements, which have promoted the progress Facebook has made in this area. Deepa Seetharaman and Jeff Horwitz report in the Wall Street Journal:
Facebook took four months to dismantle a network of accounts spreading disinformation about Ms. Franco after her murder, an event that further polarized the country ahead of the hotly contested presidential election. The company also found that a right-wing group in Brazil that supported Jair Bolsonaro, the election’s winner, was encouraging its Facebook followers to use a third-party app that enabled the group to post on their behalf twice a day. Facebook didn’t know about the activity until alerted by reporters and still couldn’t determine how widespread it was, the documents show.
The documents, shared widely among Facebook employees, show a company continuing to grapple with how its services can be manipulated to spread political disinformation during election years. They also shed light on Facebook’s preparedness for the U.S. presidential race in 2020, four years after the company was lambasted for allowing misinformation to spread across its platform and alter political discussion during a heated campaign.
Border agents are checking entrants’ Facebook and Twitter profiles — but we still don’t know how closely. (Adi Robertson / The Verge)
Twitter is becoming increasingly popular among protesters in Hong Kong — but some users are struggling to master the quirks of the platform. Who isn’t?! (Isabella Steger / Quartz)
President Trump claimed Hurricane Dorian could hit Alabama — even after weather service refuted it. The president remains one of our most powerful purveyors of misinformation. (Brian Stelter / CNN)
Meanwhile, the president’s allies are raising $2 million to investigate critical reporters, a sentence I honestly can’t believe I am typing. (Mike Allen / Politico)
The Navy has turned to popular YouTubers to help with recruitment. (Brock Vergakis / Virginian-Pilot)
A government-manded internet outage in Kashmir is wreaking havoc for residents. Instagram has proven to be a vital communication hub for one of the interviewed citizens. (Pranav Dixit / BuzzFeed)
Related: internet shutdowns are becoming more common around the world. (Patrick Kingsley / New York Times)
How Amazon’s use of contractors for delivery insulates the company from criticism when they hit pedestrians, or are exploited by the vendor. (Caroline O’Donovan and Ken Bensinger / BuzzFeed)
⭐ Jack Dorsey’s Twitter account was compromised on Friday when hackers commandeered his phone number and texted racist tweets to the Twitter shortcode. Sarah Frier and Michael Riley report at Bloomberg:
The clarification appears to support speculation that Dorsey was the victim of SIM swapping. That’s when someone convinces a mobile carrier to switch an existing number to a new SIM card they control. In this case, it may have required the hackers to have personal details that would allow them to convincingly impersonate one of Silicon Valley’s best-known figures.
More than 15 tweets, many containing obscenities and racist comments, were posted on Dorsey’s account, @jack, shortly before 4 p.m. New York time on Friday. The company started deleting the tweets from Dorsey’s verified Twitter account, which has more than 4 million followers, about 20 minutes after the messages went viral.
Twitter and Pinterest are dismantling a network of accounts who share creepy screenshots of women journalists. (Mark Di Stefano / BuzzFeed)
Facebook, like Instagram, is testing removing like counts. (Josh Constine / TechCrunch)
Facebook also ported music sharing from Instagram to its own stories product. (Ashley Carman / The Verge)
Facebook’s work on brain-reading tech sparks a call for four new privacy rights. (Sigal Samuel / Vox)
Zao is the latest deepfake face-swapping app to (1) go viral and (2) terrify everyone with its implications for (A) user privacy and (B) the integrity of our information sphere. Elsewhere, WeChat has blocked users from downloading it. (Jon Porter / The Verge)
A profile of an 8-year-old YouTuber who makes $22.5 million a year hawking toys. One thing this story does not mention: child labor laws, which parents of child YouTubers have made a convenient end run around. (Chris Stokel-Walker / FFWD)
The hottest new thing on Twitch is elderly gamers. The man behind GrndPaGaming has 200,000 subscribers, who watch him play shooters. (Kalhan Rosenblatt / NBC News)
And finally ...
Jeremy Renner is an actor who I once saw described as “like if a normal guy who works at Dicks sporting goods was granted many wishes by god.” He is also the proprietor of the official Jeremy Renner app, which people use to browse a shadow version of his Instagram feed and listen to his music. Recently the writer Stefan Heck decided to throw the world of the app into chaos by answering Renner’s question about fans’ weekend plans by announcing he (Heck) planned “to watch some porno on my personal computer.”
Due to a quirk in the app design, many Renner fans got a notification about this comment that looked as if it came from Renner itself. They have since bombed the App Store with one-star reviews:
The second thing you will find upon installing the app—if you have not already installed the official Jeremy Renner app, please feel free to take this parenthetical as an opportunity to do so—is that every push notification you receive through the app looks as though it is coming directly from Mr. Renner himself; you will also soon notice that you receive a push notification every time somebody replies to you. What this means, as you’ve probably figured out by now, is that if you post “Lookin’ good, Mr. Renner!” under a blurry video of Jeremy driving a dump truck and someone named football_jersey97 replies with “There is diarrhea shooting out of my penis,” your phone will buzz and you will receive a notification with the words “Jeremy Renner: There is diarrhea shooting out of my penis” next to a little picture of Jeremy Renner’s face.
I’ve thought about it and I’m planning to not download this app.
Talk to me
Send me tips, comments, questions, and additional R’s that YouTube should embrace: firstname.lastname@example.org.